British Journal of Anaesthesia, 116 (6): 862–9 (2016)

doi: 10.1093/bja/aew116
Respiration and the Airway

Detection of optimal PEEP for equal distribution of tidal

volume by volumetric capnography and electrical

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

impedance tomography during decreasing levels
of PEEP in post cardiac-surgery patients
P. Blankman1, A. Shono1, B. J. M. Hermans2, T. Wesselius2, D. Hasan1,3
and D. Gommers1, *
1
Department of Adult Intensive Care, Erasmus MC, Room H623, ‘s Gravendijkwal 230, Rotterdam 3015 CE,
The Netherlands, 2Institute for Biomedical Technology & Technical Medicine, University of Twente, Enschede,
The Netherlands, and 3Institute for Immunotherapy, Duderstadt, Germany
*Corresponding author. E-mail: d.gommers@erasmusmc.nl

Abstract
Background: Homogeneous ventilation is important for prevention of ventilator-induced lung injury. Electrical impedance
tomography (EIT) has been used to identify optimal PEEP by detection of homogenous ventilation in non-dependent and
dependent lung regions. We aimed to compare the ability of volumetric capnography and EIT in detecting homogenous
ventilation between these lung regions.
Methods: Fifteen mechanically-ventilated patients after cardiac surgery were studied. Ventilator settings were adjusted to
volume-controlled mode with a fixed tidal volume (Vt) of 6–8 ml kg−1 predicted body weight. Different PEEP levels were applied
(14 to 0 cm H2O, in steps of 2 cm H2O) and blood gases, Vcap and EIT were measured.
Results: Tidal impedance variation of the non-dependent region was highest at 6 cm H2O PEEP, and decreased significantly at
14 cm H2O PEEP indicating decrease in the fraction of Vt in this region. At 12 cm H2O PEEP, homogenous ventilation was seen
between both lung regions. Bohr and Enghoff dead space calculations decreased from a PEEP of 10 cm H2O. Alveolar dead space
divided by alveolar Vt decreased at PEEP levels ≤6 cm H2O. The normalized slope of phase III significantly changed at PEEP levels
≤4 cm H2O. Airway dead space was higher at higher PEEP levels and decreased at the lower PEEP levels.
Conclusions: In postoperative cardiac patients, calculated dead space agreed well with EIT to detect the optimal PEEP for an
equal distribution of inspired volume, amongst non-dependent and dependent lung regions. Airway dead space reduces at
decreasing PEEP levels.

Key words: capnography; mechanical ventilation; peep; ventilator induced lung injury

Accepted: March 17, 2016

© The Author 2016. Published by Oxford University Press on behalf of the British Journal of Anaesthesia.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/
licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
For commercial re-use, please contact journals.permissions@oup.com

862
 Optimal PEEP detection by volumetric capnography and EIT | 863

(PCV).22 23 In addition, EIT has been used to visualize hyperinfla-

Editor’s key points
tion in the non-dependent region at a certain PEEP level when
• This pilot study compared volumetric capnography and pixel-compliance is decreased.24 25
electrical impedance tomography (EIT) in detecting distri- Suter and colleagues26 defined optimal PEEP as the balance of
bution of ventilation after cardiac surgery. adequate PaO2 levels, good compliance and elimination of CO2.
• EIT values approximated to calculated dead space and de- However, we believe that stress and strain are important contribu-
tected optimal PEEP to equalize gas distribution in dependent tors to lung injury, as both factors increase with inhomogeneous
and non-dependent lung regions. ventilation.5 6 Therefore, the main goal of this study was to com-
• At 12 cm H20 PEEP, ventilation was distributed homoge- pare the results of Vcap measurements with that of EIT in finding
neously between dependent and non-dependent regions. the optimal PEEP, with an equal distribution of the inspiratory
• However, excessive PEEP applied to healthy lungs caused air- volume among the dependent and the non-dependent regions.
way distension rather than improving alveolar ventilation.

Methods

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

Study population
Ventilator-induced lung injury (VILI) is a well-recognized compli-
cation of mechanical ventilation, which occurs in both patients In this pilot-study, 15 mechanically ventilated patients, who had
with acute respiratory distress syndrome (ARDS)1 and healthy undergone coronary-artery bypass grafting and/or cardiac-valve
lungs.2 3 The deleterious effects of mechanical ventilation are surgery, admitted to the cardiothoracic intensive care unit (ICU)
caused by large amounts of stress and strain acting on lung tis- were included. The local medical ethical committee approved the
sue, as a result of inhomogeneous ventilation of the lungs.4–6 study protocol and written informed consent was obtained from
Although, protective lung strategies using low tidal volumes each patient or their relatives. Data were collected between
have been strongly recommended to prevent the development January and July 2014.
of VILI, the amount of PEEP that should be applied is still under The inclusion criteria were age >18 yrs, written informed con-
debate.7 8 The positive effect of PEEP depends on the recruitability sent, haemodynamically stable. Exclusion criteria were: presence
of lung tissue, which varies between patients,9 10 but also from of a cardiac pacemaker, pneumothorax, thoracic deformations
day to day. Therefore, setting the PEEP level without a reliable and severe airflow limitation (defined as forced expiratory
tool to estimate the distribution of inspiratory tidal volume (Vt) volume in 1 s below 70% of forced vital capacity).
at the bedside is a challenging activity.
Volumetric capnography (Vcap) is a non-invasive technique
Study protocol
that describes the CO2 exhalation during one breath, and can be
used to calculate alveolar and airway dead space.11 Because elim- For Vcap measurements a mainstream CO2 sensor was placed
ination of CO2 depends on alveolar ventilation and pulmonary between the tracheal tube and ventilator tubing’s, and was
perfusion, V/Q mismatches in either hyperinflated or collapsed connected to a NICO-capnograph (Novametrix, Wallinford, Con-
lung areas will affect the amount of exhaled CO2 and thereby necticut, USA). Specific software (Analysis plus, Novametrix,
Vcap. Recently, a multi-centre observational study showed that Wallinford, Connecticut, USA) was used to record all Vcap data.
an increased dead space fraction in patients with ARDS was asso- In order to perform electrical impedance tomography (EIT) mea-
ciated with an increased mortality.12 Several studies showed the surements, a silicon belt with 16 electrodes was placed around
capability of Vcap in detecting optimal PEEP level during mech- the thoracic cage between the 5th and 6th intercostal space (Pul-
anical ventilation in various lung pathological conditions.13–15 movista 500, Dräger, Lübeck, Germany). Ventilator settings were
Böhm and colleagues14 showed in 11 morbidly obese patients set to volume-controlled mode (Evita-infinity, Dräger, Lübeck,
undergoing bariatric surgery, Slope III (SIII) has a high sensitivity Germany), with a fixed tidal volume of 6–8 ml kg−1 predicted
and specificity to detect lung recruitment. They concluded that body weight and inspiration/expiration (I/E) ratio of 1:2. We delib-
SIII was useful for identifying appropriate levels of PEEP in those erately choose volume control in order to avoid a change in the
patients. In 20 obese patients undergoing laparoscopic bariatric min volume because of differences in lung compliance at differ-
surgery Tusman and colleagues13 found that both Vcap and ent levels of PEEP. The initial setting of the respiratory rate and
pulse-oximetry, as compared with respiratory compliance, are fraction of inspired oxygen was adjusted to maintain an end-
accurate parameters for the detection of alveolar collapse. Feng- tidal carbon dioxide tension and oxygen saturation within a
mei and colleagues15 found that best PEEP based on dead space range of 35–45 mm Hg and 97–100%, respectively. PEEP was set
fraction corresponded well with best PEEP according to highest according to the attending physician. Vt, respiratory rate (RR), I/E
respiratory compliance, in 23 ARDS patients. Therefore, Vcap is ratio remained unchanged throughout the entire study period.
a promising technique to detect alveolar collapse and recruit- We are reluctant to apply high levels of PEEP (>15 cm H2O) in
ment at the bedside. postoperative patients after cardiac surgery who obviously have
Electrical Impedance Tomography (EIT) is a non-invasive, no ALI or ARDS, as these patients are generally prone to haemo-
radiation-free, real time imaging modality, which has proved to dynamic instability. Therefore, the patients were not subjected to
correlate well with Computed Tomography (CT), according to as- recruitment manoeuvers. After baseline measurements of Vcap
sessment of changes in gas volume and tidal volume.16–18 and EIT, the PEEP level was increased to 14 cm H2O. This PEEP
Recently, we showed that the intratidal gas distribution calculated level was maintained for 15–20 min in order to reach a steady
from EIT measurements, is able to define a patient specific PEEP state situation, as assessed by a stable signal of the volume of ex-
level, at which the lungs are homogeneously ventilated among haled carbon dioxide (VCO2). Thereafter, the PEEP was decreased
dependent and non-dependent lung regions. This parameter from 14 to 0 cm H2O PEEP in steps of 2 cm H2O. Each PEEP level was
was evaluated in both experimental studies and in patients,19–21 applied for five to 10 min depending on haemodynamically
but also during pressure support ventilation (PSV), Neurally As- stability. Blood gas analysis, Vcap and EIT were measured during
sisted Ventilatory Assist (NAVA) and pressure control ventilation each PEEP step.
864 | Blankman et al.

Data analysis normalize the value and make it comparable between patients.28 29
Normalized Slope III (SnIII) represents the homogeneity of ventila-
Volumetric capnography; model fitting and parameters
tion and pulmonary perfusion, which is considered as a good indi-
Vcap, which is constructed from expired CO2 concentration and
cator of the global V/Q matching. SnIII increases in lung conditions
expired volume, can be represented by a mathematical function,
associated with a mismatch of ventilation and perfusion such as
which is previously introduced by Tusman and colleagues.27 28
atelectasis or pulmonary artery embolism, whereas low values
This makes the analysis of the Vcap data less susceptible to
indicate homogeneous V/Q matching.
noise. To obtain this mathematical function from Vcap data, a
Physiological dead space was calculated according to the for-
non-linear least square curve fitting algorithm according to
mula developed by Bohr, and the Enghoff modification of Bohr’s
Tusman and colleagues was used in a custom-made Matlab®
formula30 (formula 1+formula 2):
(The MathWorks, Natick, MA, USA) program. From this model,
(Mean alveolar partial pressure of CO2 ðPA CO2 Þ; mixed expired
parameters were derived for further analysis. First, the Vcap  2 ¼Þ)
partial pressure of CO2 ðPECO
curve was divided into three phases (Phase I, II and III) (Fig. 1).
Phase I indicates the CO2 volume originated from the airways,
 2
PA CO2  PECO

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

whereas phase III is the CO2 from the alveoli. Phase II therefore VDBohr =VT ¼ ð1Þ
is the phase with CO2 from both the airways and alveoli. The in- PA CO2
flection point of Phase II (Point A), which is determined by the
maximum of the first derivative of the fitted model, was defined  2
PaCO2  PECO
as the mean airway-alveolar interface that separates the airway VDEnghoff =VT = ð2Þ
PaCO2
from the alveolar compartment. Airway dead space (VDaw) is
the volume from the beginning of the expiration until point Alveolar dead space (VDalv) was computed by subtracting VDaw
A. Slope II (SII) is the slope of the inflection point of Phase II. from VDEnghoff. We chose to subtract VDaw from VDEnghoff as the
Slope III (SIII) is the slope of phase III, which is calculated accord- VDEnghoff contains all causes of V/Q inhomogeneity, whereas
ing to Tusman and colleagues.27 28 Therefore, phaseIII was divided the Bohr formula is not supposed to contain shunts and low
into three even sized segments. A line was fitted through the V/Q regions. The ratio of VDalv to alveolar tidal volume (VTalv)
second segment of phaseIII by the least-square method. The was obtained by dividing VDalv by VTalv (VTalv=VT−VDaw), in
slope of this fitted line was determined as SIII. Thereafter slope which VTalv represents alveolar tidal volume.
III was divided by the fraction of end tidal CO2 (FECO2) in order to
EIT parameters
EIT data were recorded with a sample frequency of 20 Hz. Data
was reconstructed and analysed using special software (EITdiag,
Phase I Phase II Phase III Dräger, Lübeck, Germany). For each PEEP level, 10 to 20 consecu-
tive breaths at the end of each PEEP step were selected for ana-
PeCO2 (mm Hg)

PaCO2 lysis, with the assumption that the lungs were best adapted to
the ventilator settings at that time. EIT data contains signals in-
duced by the respiratory system and the circulatory system.
PACO2 Therefore, to filter out signals related to the circulatory system,
EIT data was filtered using a low pass-filter set to 40 beats min-1.
The reconstructed image was represented as a ventilation distri-
bution map, which was generated by tidal impedance changes
PECO2 caused, by inspiration and expiration. These impedance changes
are presented as tidal impedance variation (TIV) and have been
A
shown to correlate well with gas volume changes, as measured
by CT-scans.16–18 The surface area of the distribution map was
divided into two equal regions of interest (ROIs), to know the
non-dependent and dependent lung regions (Supplementary
data, Fig. S1). The surface area of the EIT map was kept equal
for all PEEP steps.
Expired volume (ml)
The intratidal gas distribution (ITV) was developed by Löwha-
Vdaw Vtalv gen, and colleagues21 (formula 3). The ITV describes the amount
of the impedance distributed to each region of interest within one
inspiration. For this calculation Löwhagen and colleagues21
Vt
divided the inspiratory part of the global impedance curve into
Phase I: airway CO2 eight equal volume sections. In other words each part of the
Phase II: mixed airway and alveolar CO2 inspiration is 12.5% of the entire inspiration. Thereafter, they
Phase III: alveolar CO2
transposed the time needed for each section of the inspiration
to the regional curves. In this way they were able to calculate
Fig 1 Schematic model of three phases of the expiration. The Vcap curve is the contribution of four different regions of interest to the
divided in three phases. Phase I represents exhaled CO2 from the airways, inspiration, during a single breath. According to our previous
whereas phase III represents exhaled CO2 from the alveoli. Phase II is a publications19 20 22 31 we calculated the ITV for two instead of
mixed phase with CO2 from both the airways and alveoli. Point A is the
four regions of interest. Using the ITV calculation one is able to
inflection point of phase II, which is the theoretical point where the
analyse the ventilation homogeneity during the inspiration. In
airway compartment is separated from the alveolar compartment. VDaw,
airway dead space, Vtalv, alveolar tidal volume, Vt, tidal volume. order to reliably calculate the ITV, the filtered EIT signals were
re-sampled with a frame rate of 40 Hz. This step is important in
 Optimal PEEP detection by volumetric capnography and EIT | 865

order to divide the inspiratory part of the global TIV curve reliably both regions can be more or less than 100%, as compared with
into eight iso-volume sections (ITV, Intratidal Gas Distribution; TIV, PEEP 14 cm H2O. At 14 cm H2O of PEEP, most of the ventilation
Tidal Impedance Variation; ROI, Region of Interest; t, iso-volume part).21 was distributed to the dependent lung region and from PEEP
level of 10 cm H2O and lower, the non-dependent lung region be-
came predominant ventilated (Fig. 2). TIV in the dependent lung
ITV18 TIVROI
Fractional regional ITV18 ¼ ð3Þ region decreased during each PEEP step reduction from a PEEP of
ITV18 TIVGlobal
12 cm H2O, indicating less ventilation at lower PEEP levels as a re-
In addition to the capnography and EIT parameters we also sult of collapse (Fig. 2). TIV in the non-dependent region had the
calculated the dynamic compliance. Dynamic compliance was highest value at 6 cm H2O of PEEP (Fig. 2). In addition, ventilation
calculated by dividing the expiratory tidal volume by Peak was evenly distribution among dependent and non-dependent
Inspiratory Pressure (PIP) minus PEEP. regions at 12 and 10 cmH2O PEEP (Fig. 3).
Table 2 shows changes in different Vcap parameters. VDBohr/
Statistical analysis VT and VDEnghoff/VT significantly decreased from a PEEP of 10 cm
H2O and less as compared with 14 cm H2O PEEP. The ratio of al-

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

Statistical analyses were carried out using SPSS 21 (Chicago, IL, veolar dead space to alveolar tidal volume significantly decreased
USA). Data are presented as mean () unless otherwise specified. from a PEEP level of 6 cm H2O and less as compared with 14 cm
Normal distribution of the data was tested using the Kolmogorov- H2O PEEP. The normalized slope of phase III significantly changed
Smirnov test, whereas homoscedasticity was tested using the at a PEEP levels of 4 cm H2O or less, whereas volume of phase III
Brown-Forsythe test. Changes in Vcap and EIT parameters during corrected for tidal volume (VIII/VT) increased significantly, as
the entire protocol were tested using mixed linear models. Differ- compared with 14 cm H2O PEEP at ≤8 cm H2O PEEP, indicating
ences between two sequential PEEP steps were tested using the more alveolar ventilation at lower PEEP levels. In addition, VDaw
paired Student’s t-test for normal distributed data, and using decreased significantly from a PEEP level of 8 cm H2O or less.
the Wilcoxon rank sum test for not normal distributed data. For Shunt fraction calculated as the difference between arterial CO2
all statistical test P<0.05 was considered statistically significant. minus end-tidal CO2 (Pa-ETCO2)32 only significantly differed at
ZEEP.
Results Figure 4 shows the effect of PEEP on the Vcap curve. When
lower PEEP levels are applied the Vcap curve is shifted to the
The initial physiological data, measured shortly before 14 cm H2O
left, as a consequence of decreased VDaw. VDaw/VT significantly
was applied, is summarized in Table 1. During the different PEEP
decreased at 10 cm H2O and lower as compared with 14 cm H2O
levels, blood gases and dynamic compliance remained stable,
PEEP (Table 2).
down to a PEEP level of 2 cm H2O and decreased significantly at
0 PEEP (ZEEP) (Table 2).
Figure 2 shows the TIV for the dependent and non-dependent
lung region. The 14 cm H2O PEEP step was chosen as reference,
Discussion
and the TIV of both lung regions together was set to 100% at In this study we found that dead space calculations according to
14 cm H2O, Therefore, at lower PEEP levels the sum of TIV of Bohr and Enghoff agreed well with the PEEP level, resulting in

Table 1 Data are presented as mean () unless otherwise specified. Predicted body weight (PBW), Cardiopulmonary bypass (CPB), Coronary
artery by-pass graft (CABG), Arterial partial pressure of O2 ðPaO2 Þ, Fraction of inspired oxygen ðFIO2 Þ, Arterial partial pressure of CO2 ðPaCO2 Þ

Patient characteristics
No. of patients 15
Age (yr) 70 (8)
Male/Female 13/2
Height (cm) 176 (13)
Weight (kg) 86 (18)
Predicted body weight (kg) 70 (13)
BMI 28 (4)
CPB time (min) 112 (35)
Valve CABG+Valve
Type of surgery CABG
replacement replacement
10 3 2
Haemodynamic data at ICU admission
Mean arterial pressure (mm Hg) 70 (9)
Heart rate (BPM) 77 (15)
Ventilator settings and respiratory measurements at ICU admission
Positive end-expiratory pressure (cm H2O) 8 (1)
Peak inspiratory pressure (cm H2O) 22 (2)
Expiratory tidal volume (ml) 494 (74)
Expiratory tidal volume/Predicted body weight (ml kg PBW−1) 7.1 (0.6)
Respiratory rate(Bpm) 17 (2)
PaO2 =FIO2 (mm Hg) 350 (101)
PaCO2 (mm Hg) 42.2 (4.5)
866 | Blankman et al.

Table 2 Data are presented as mean (). The first statistical change compared with 14 cm H2O PEEP is indicated by *P<0.05 was considered
significant. Tidal volume (VT), Airway dead space (VDaw), Alveolar dead space (VDalv), Alveolar dead space to alveolar tidal volume ratio
(VDalv/VTalv), Normalized slope of phase III (SnIII), Amount of expired CO2 within one breath (VTCO2,br), Volume of phase III to tidal volume
ratio (VIII/VT), Arterial minus end-tidal partial pressure of CO2 (Pa-ETCO2), Arterial partial pressure of O2 ðPaO2 Þ, Arterial partial pressure of
CO2 ðPaCO2 Þ, Fraction of inspired oxygen ðFIO2 Þ

Dead space variables, blood gas analysis and compliance during the decremental PEEP trial
PEEP (cm H2O) 14 12 10 8 6 4 2 0

VT (ml) 472 (80) 476 (77) 474 (80) 473 (77) 472 (76) 471 (78) 474 (81) 473 (81)
VDaw (ml) 214 (40) 209 (42) 204 (39) 196 (40)* 188 (39) 182 (36) 174 (32) 167 (26)
VDalv (ml) 52 (37) 52 (34) 47 (31) 45 (33) 42 (33)* 35 (32) 39 (31) 35 (29)
VDBohr/VT 0.48 (0.06) 0.47 (0.07) 0.45 (0.07)* 0.43 (0.07) 0.40 (0.07) 0.38 (0.07) 0.35 (0.07) 0.33 (0.07)
VDEnghoff/VT 0.57 (0.07) 0.55 (0.08) 0.53 (0.08)* 0.52 (0.08) 0.49 (0.08) 0.47 (0.09) 0.46 (0.08) 0.44 (0.08)

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

VDalv/VTalv 0.19 (0.11) 0.18 (0.1) 0.17 (0.09) 0.15 (0.09) 0.14 (0.1)* 0.12 (0.1) 0.13 (0.09) 0.11 (0.09)
SnIII (mm Hg ml−1) 0.95 (1.19) 0.73 (0.64) 0.62 (0.31) 0.62 (0.43) 0.58 (0.33) 0.48 (0.25)* 0.47 (0.26) 0.50 (0.21)
VTCO2,br (ml) 12.5 (3.3) 13.3 (3.7)* 13.9 (3.7) 14.3 (3.6) 14.7 (3.7) 15.2 (3.7) 15.7 (3.8) 16.0 (3.9)
VIII/VT 0.42 (0.14) 0.39 (0.13) 0.41 (0.14) 0.43 (0.13)* 0.46 (0.13) 0.46 (0.13) 0.48 (0.12) 0.51 (0.11)
Pa-ETCO2 (mm Hg) 3.6 (2.8) 3.4 (2.5) 3.7 (2.6) 3.2 (2.5) 2.9 (2.7) 2.7 (2.4) 3.1 (2.4) 2.8 (2.5)*
PaCO2 (mm Hg) 43 (5) 43 (5) 44 (5) 43 (5) 42 (5) 42 (5) 42 (5) 41 (5)*
PaO2 (mm Hg) 138 (35) 140 (32) 145 (25) 145 (24) 145 (23) 139 (21) 132 (24) 127 (23)*
PaO2 =FIO2 (mm Hg) 346 (87) 349 (80) 359 (71) 371 (71) 372 (68) 357 (65) 338 (72) 324 (70)*
Dynamic compliance 34 (6) 36 (8) 35 (6) 35 (7) 35 (7) 34 (7) 33 (7) 31 (7)*
(ml cm H2O−1)

with the non-dependent lung region and found that the depend-
ent lung region had the highest TIV at 12 cm H2O PEEP and de-
Tidal ventilation distribution at different PEEP levels
creased with each decremental PEEP step, indicating collapse in
100 that region. At a PEEP level of 12 cm H2O the dependent and
Mean tidal impedance variation (%)

* non-dependent lung regions were in balance. Above this PEEP

*
80 the dependent region received most of the tidal volume, whereas
below this PEEP the non-dependent region was predominantly
ventilated. Thus EIT is the first device able to visualize collapse
60
and hyperinflation of lung tissue at the bedside.
In an experimental study by Tusman and colleagues32 a RM
40 followed by decremental PEEP trial from 24 to 0 cm H2O in steps
of 2 cm H2O was performed, in eight lung lavaged pigs. They in-
20 vestigated the ability of Vcap to detect optimal PEEP, as compared
with computed tomography scans (CT), and found that dynamic
† † † † †
compliance, VDalv/VTalv and Pa-ETCO2 correlated best with the
0
amount of non-aerated lung tissue as detected by CT-scans.
14 12 10 8 6 4 2 0 Yang and colleagues33 performed a decremental PEEP trial on
PEEP (cm H2O) eight lung-lavaged piglets. After a RM they stepwisely reduced
the PEEP from 20 cm H2O to 4 cm H2O in steps of 4 cm H2O. At
PEEP levels below 16 cm H2O they found that VDalv/VTalv was
Fig 2 Tidal Impedance Variation at different PEEP levels. Data are shown as
well correlated to non-aerated and normally aerated lung tissue
mean (). In the dependent lung region the ventilation distribution was
decreased when PEEP was lowered, whereas the non-dependent region as assessed by CT-scans. In addition, they found that the lowest
received more ventilation as compared with PEEP 14 cm H2O. * Indicates a SIII value was closely related to best lung function. They con-
significant reduction in TIV of the non-dependent region according to 6 cluded that VDalv/VTalv and SIII are best predictors for alveolar col-
cm H2O. † Indicates a significant reduction in TIV of the dependent region lapse or hyperinflation. Therefore, from these two studies it can
as compared with 12 cm H2O. Dashed lines represents the interpolation
be concluded that Vcap is a reliable tool to detect both collapse
lines; open circles=non-dependent region; solid circles=dependent region.
and hyperinflation of lung tissue, as confirmed by CT-scans,
P<0.05 was considered significant.
which is the golden standard to assess alveolar collapse or
hyperinflation.
Maisch and colleagues34 performed an incremental and de-
homogeneous ventilation as measured by EIT. In contrast, cremental PEEP trial in 20 anesthetized patients with healthy
VDalv/VTalv and SnIII were insensitive to detect lung inhomogen- lungs, undergoing faciomaxillary surgery. During the PEEP trial,
eity. In addition, we found that in relatively healthy lungs, PEEP levels between 0 and 15 cm H2O were applied in PEEP
without the application of a RM, higher PEEP levels mainly induce steps of 5 cm H2O. They showed that after a RM physiological
airway distention rather than improve alveolar ventilation. dead space was the lowest at 10 cm H2O PEEP, whereas static com-
Recently, we demonstrated that homogeneous ventilation by pliance of the respiratory system was the highest at that PEEP
means of ITV agreed well with a PEEP level, with the highest level. In addition, functional residual capacity and PaO2 were in-
dynamic compliance.20 We compared the whole dependent sensitive parameters for detection of alveolar hyperinflation as
 Optimal PEEP detection by volumetric capnography and EIT | 867

PEEP 14 PEEP 12 PEEP 10 PEEP 8

100

80
Mean proportion of contribution to the fractions of mean TIV (%)

60

40

20

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

0

12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100

PEEP 6 PEEP 4 PEEP 2 PEEP 0

100

80

60

40

20

12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100 12.5 25 37.5 50 62.5 75 87.5 100

Proportion of inspiration (%)

Fig 3 Mean intratidal gas distribution (ITV) curve of all the patients at each PEEP step. Data are shown as mean (). ITV curve represents the mean percentile
contribution (%) of ventilation distribution in non-dependent and dependent lung regions during the entire inspiration. At PEEP 12 and 10 cm H2O, intratidal gas
distribution curves stayed close to each other, showing equal distribution to both regions. Dashed lines represents the interpolation lines; open circles=non-
dependent region; solid circles=dependent region.

they decreased with each PEEP step. Therefore, they concluded

that physiological dead space and static compliance of the
respiratory system, are suitable parameters for PEEP titration.
Interestingly, we found that the PEEP level at which VDBohr/VT
and VDEnghoff/VT significantly decreased (Table 2), suggesting a
reduction in V/Q mismatch, agreed well with the PEEP level at
PeCO2 (mm Hg)

which the inspired air was homogeneously distributed to the

dependent and non-dependent lung regions, as measured by
EIT (Figs 2 and 3). However, at lower PEEP levels, VDBohr/VT and
VDEnghoff/VT keeps decreasing indicating that this parameter
adds no additional information about ventilation homogeneity
at lower PEEP. In addition, the exhaled CO2 per breath (VCO2,br)
PEEP 0 cm H2O and the SnIII both showed that elimination of CO2 improved at
lower PEEP levels (Table 2), but no optimum could be found.
PEEP 14 cm H2O
Therefore, in the present study design and study population,
A Vcap is not able to detect homogeneous ventilation.
Our finding in Vcap that PEEP mainly increases the airway-
Expired volume (ml)
volume rather than improves alveolar ventilation is apparently
in contradiction with the results of other studies published earl-
Fig 4 Effect of increased PEEP on the Vcap curve. Figure 4 demonstrates the ier.14 32 33 Lower PEEP levels cause a shift of the Vcap curve to the
effect of two PEEP levels on the Vcap curve. As a result of the PEEP
left (Fig. 4), indicating increased airway dead space at higher
application the Vcap curve shifts to the right, indicating an increase in
PEEP levels. These opposing results could be explained by the
airway dead space. PeCO2 = Expiratory carbondioxide pressure; Vcap =
volumetric capnography. fact that post-cardiac surgery patients have relatively healthy
lungs. Although the patients in our study may have atelectasis,
868 | Blankman et al.

there is no question of surfactant depletion and we assume that Authors’ contributions

the atelectasis is recruited without the need for high PEEP levels.
Study design/planning: P.B., D.G.
In contrast, the previous studies used the lung-lavaged models
Study conduct: P.B., A.S., B.H., T.W.
with large amount of shunt area, which are recruitable at high
Data analysis: P.B., A.S., B.H., T.W., D.H., D.G.
PEEP levels. However, Nieman and colleagues35–41 found during
Writing paper: P.B., A.S., B.H., T.W., D.H., D.G.
in vivo microscopy in normal and surfactant deactivated pigs,
Revising paper: all authors
during ventilation with tidal volumes of 6, 12 and 15 ml kg−1,
that in normal lungs the alveolar area at end-inspiration and
end-expiration (I-EΔ) did not differ between the three different Supplementary material
tidal volumes, whereas in the surfactant deactivated lungs I-EΔ Supplementary material is available at British Journal of Anaesthesia
increased significantly at higher tidal volumes. In a next online.
study35 they used different tidal volumes (6, 12, 15 ml kg−1) and
PEEP levels (5, 10, 20 cm H2O) and found that I-EΔ did not change
Declaration of interest
in normal lungs, whereas in surfactant deactivated lungs the al-

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

veolar size significantly increased, as compared with normal None declared.
lungs with equal settings. From these studies it can be concluded
that in normal lungs with stable alveoli, PEEP volume results in Funding
airway distention, whereas in ALI/ARDS lungs, PEEP keeps the
Department of Adult Intensive Care, Erasmus MC, Rotterdam,
alveoli open and stabilized. In addition, airway distention as a
The Netherlands.
result of high PEEP might also explain why physiological dead
space calculated according to Enghoff, increases at higher PEEP
levels, whereas PaCO2 did not significantly differ during the differ-
References
ent PEEP levels, except at ZEEP. This explains the absence of 1. Plotz FB, Slutsky AS, van Vught AJ, Heijnen CJ. Ventilator-
a change in SIII and the left shift of SII in the Vcap curves at induced lung injury and multiple system organ failure: a crit-
decreasing PEEP levels in our patients. ical review of facts and hypotheses. Intensive Care Med 2004;
There are limitations to this study. Firstly, we did not apply a 30: 1865–72
RM before the PEEP reduction, in order to avoid serious deterior- 2. Futier E, Constantin JM, Paugam-Burtz C, et al. A trial of intrao-
ation of haemodynamics in the direct postoperative period. It has perative low-tidal-volume ventilation in abdominal surgery.
been shown that Vcap parameters after an RM are more sensitive N Engl J Med 2013; 369: 428–37
to detect changes in CO2 elimination, as compared with PEEP 3. Schultz MJ, Haitsma JJ, Slutsky AS, Gajic O. What tidal vo-
trials without an RM.42 43 Secondly, EIT describes the ventilation lumes should be used in patients without acute lung injury?
distribution in a lung slice of approximately 5–10 cm thick.42 44 Anesthesiology 2007; 106: 1226–31
Therefore, the image obtained from the single cross-section 4. Mead J, Takishima T, Leith D. Stress distribution in lungs: a
does not represent the entire lung. In addition, PEEP increases model of pulmonary elasticity. J Appl Physiol 1970; 28: 596–608
results in changed chest size and intrathoracic blood volume. 5. Protti A, Cressoni M, Santini A, et al. Lung stress and strain
However, by using two large regions of interest, the correspond- during mechanical ventilation: any safe threshold? Am
ing lung tissue to the individual pixel is limited. The effect J Respir Crit Care Med 2011; 183: 1354–62
of changed intrathoracic blood volume is limited, as EIT is a hun- 6. Protti A, Andreis DT, Monti M, et al. Lung Stress and Strain
dred times more sensitive to impedance changes caused by During Mechanical Ventilation: Any Difference Between
aeration, as compared with impedance changes caused by the Statics and Dynamics?. Crit Care Med 2013; 41: 1046–55
cardiac system.45–47 Thirdly, the time intervals of 5–10 min at 7. Briel M, Meade M, Mercat A, et al. Higher vs lower positive
each PEEP level might be too short to achieve a steady state for end-expiratory pressure in patients with acute lung injury
obtaining reliable data. However, a recent study showed that and acute respiratory distress syndrome: systematic review
elimination of CO2 reached a new stable period within five min and meta-analysis. JAMA 2010; 303: 865–73
of a PEEP change.48 It is known that Vcap is influenced by both 8. Putensen C, Theuerkauf N, Zinserling J, Wrigge H, Pelosi P.
the cardiac system and the pulmonary system. In this study we Meta-analysis: ventilation strategies and outcomes of the
did not perform cardiac output measurements. Therefore we acute respiratory distress syndrome and acute lung injury.
cannot assess Vcap changes induced by reduced cardiac output Ann Intern Med 2009; 151: 566–76
at higher PEEP levels. However, we found no significant differ- 9. Cressoni M, Cadringher P, Chiurazzi C, et al. Lung inhomogen-
ences in bp and heart rate between PEEP 14 and ZEEP. Therefore, eity in patients with acute respiratory distress syndrome. Am
we assume that the influence of the change in cardiac output J Respir Crit Care Med 2014; 189: 149–58
with different PEEP levels on dead space was not profound. 10. Gattinoni L, Caironi P, Cressoni M, et al. Lung recruitment
in patients with the acute respiratory distress syndrome.
N Engl J Med 2006; 354: 1775–86
Conclusion 11. Suarez-Sipmann F, Bohm SH, Tusman G. Volumetric capno-
In postoperative cardiac patients with low shunt fractions, dead graphy: the time has come. Curr Opin Crit Care 2014; 20: 333–9
space calculations according to Bohr and Enghoff agreed well 12. Kallet RH, Zhuo H, Liu KD, Calfee CS, Matthay MA. The
with EIT. EIT is able to detect the optimal PEEP level for equal Association Between Physiologic Dead-Space Fraction and
distribution of inspired air, among the non-dependent and the Mortality in Subjects With ARDS Enrolled in a Prospective
dependent regions of the lungs. In contrast, VDalv/VTalv and Multi-Center Clinical Trial. Respir Care 2014; 59: 1611–8
SnIII were unable to detect this distribution. In addition, we 13. Tusman G, Groisman I, Fiolo FE, et al. Noninvasive monitoring
found that in relatively healthy lungs, applying higher PEEP levels of lung recruitment maneuvers in morbidly obese patients:
mainly induced airway distention rather than improvement of the role of pulse oximetry and volumetric capnography.
alveolar ventilation. Anesth Analg 2014; 118: 137–44
 Optimal PEEP detection by volumetric capnography and EIT | 869

14. Bohm SH, Maisch S, von SA, et al. The effects of lung recruit- 31. Blankman P, van der Kreeft SM, Gommers D. Tidal ventilation
ment on the Phase III slope of volumetric capnography in distribution during pressure-controlled ventilation and
morbidly obese patients. Anesth Analg 2009; 109: 151–9 pressure support ventilation in post-cardiac surgery patients.
15. Fengmei G, Jin C, Songqiao L, Congshan Y, Yi Y. Dead space Acta Anaesthesiol Scand 2014; 58: 997–1006
fraction changes during PEEP titration following lung recruit- 32. Tusman G, Suarez-Sipmann F, Bohm SH, et al. Monitoring
ment in patients with ARDS. Respir Care 2012; 57: 1578–85 dead space during recruitment and PEEP titration in an
16. Frerichs I, Hinz J, Herrmann P, et al. Detection of local lung air experimental model. Intensive Care Med 2006; 32: 1863–71
content by electrical impedance tomography compared with 33. Yang Y, Huang Y, Tang R, et al. Optimization of positive end-
electron beam CT. J Appl Physiol 2002; 93: 660–6 expiratory pressure by volumetric capnography variables in
17. Meier T, Luepschen H, Karsten J, et al. Assessment of regional lavage-induced acute lung injury. Respiration 2014; 87: 75–83
lung recruitment and derecruitment during a PEEP trial based 34. Maisch S, Reissmann H, Fuellekrug B, et al. Compliance and
on electrical impedance tomography. Intensive Care Med 2008; dead space fraction indicate an optimal level of positive
34: 543–50 end-expiratory pressure after recruitment in anesthetized
18. Victorino JA, Borges JB, Okamoto VN, et al. Imbalances in re- patients. Anesth Analg 2008; 106: 175–81, table

Downloaded from https://academic.oup.com/bja/article/116/6/862/2566524 by guest on 03 May 2024

gional lung ventilation: a validation study on electrical imped- 35. Halter JM, Steinberg JM, Gatto LA, et al. Effect of positive end-
ance tomography. Am J Respir Crit Care Med 2004; 169: 791–800 expiratory pressure and tidal volume on lung injury induced
19. Bikker IG, Blankman P, Specht P, Bakker J, Gommers D. Global by alveolar instability. Crit Care 2007; 11: R20
and regional parameters to visualize the ‘best’ PEEP during a 36. Nieman GF, Bredenberg CE, Clark WR, West NR. Alveolar
PEEP trial in a porcine model with and without acute lung function following surfactant deactivation. J Appl Physiol
injury. Minerva Anestesiol 2013; 79: 983–92 Respir Environ Exerc Physiol 1981; 51: 895–904
20. Blankman P, Hasan D, Groot JE, Gommers D. Detection of 37. Schiller HJ, McCann UG, Carney DE, Gatto LA, Steinberg JM,
‘best’ positive end-expiratory pressure derived from electric- Nieman GF. Altered alveolar mechanics in the acutely injured
al impedance tomography parameters during a decremental lung. Crit Care Med 2001; 29: 1049–55
positive end-expiratory pressure trial. Crit Care 2014; 18: R95 38. Halter JM, Steinberg JM, Schiller HJ, et al. Positive end-
21. Lowhagen K, Lundin S, Stenqvist O. Regional intratidal gas expiratory pressure after a recruitment maneuver prevents
distribution in acute lung injury and acute respiratory dis- both alveolar collapse and recruitment/derecruitment. Am J
tress syndrome–assessed by electric impedance tomography. Respir Crit Care Med 2003; 167: 1620–6
Minerva Anestesiol 2010; 76: 1024–35 39. Schiller HJ, Steinberg J, Halter J, et al. Alveolar inflation during
22. Blankman P, Hasan D, van Mourik MS, Gommers D. Ventila- generation of a quasi-static pressure/volume curve in the
tion distribution measured with EIT at varying levels of acutely injured lung. Crit Care Med 2003; 31: 1126–33
pressure support and Neurally Adjusted Ventilatory Assist 40. Steinberg JM, Schiller HJ, Halter JM, et al. Alveolar instability
in patients with ALI. Intensive Care Med 2013; 39: 1057–62 causes early ventilator-induced lung injury independent of
23. Blankman P, van der Kreeft SM, Gommers D. Tidal ventilation neutrophils. Am J Respir Crit Care Med 2004; 169: 57–63
distribution during pressure-controlled ventilation and 41. Steinberg J, Schiller HJ, Halter JM, et al. Tidal volume increases
pressure support ventilation in post-cardiac surgery patients. do not affect alveolar mechanics in normal lung but cause
Acta Anaesthesiol Scand 2014; 58: 997–1006 alveolar overdistension and exacerbate alveolar instability
24. Bikker IG, Leonhardt S, Reis MD, Bakker J, Gommers D. Bed- after surfactant deactivation. Crit Care Med 2002; 30: 2675–83
side measurement of changes in lung impedance to monitor 42. Erlandsson K, Odenstedt H, Lundin S, Stenqvist O. Positive
alveolar ventilation in dependent and non-dependent parts end-expiratory pressure optimization using electric imped-
by electrical impedance tomography during a positive end- ance tomography in morbidly obese patients during laparo-
expiratory pressure trial in mechanically ventilated intensive scopic gastric bypass surgery. Acta Anaesthesiol Scand 2006;
care unit patients. Crit Care 2010; 14: R100 50: 833–9
25. Costa EL, Borges JB, Melo A, et al. Bedside estimation of 43. Tusman G, Bohm SH, Suarez-Sipmann F, Turchetto E.
recruitable alveolar collapse and hyperdistension by electric- Alveolar recruitment improves ventilatory efficiency of the
al impedance tomography. Intensive Care Med 2009; 35: 1132–7 lungs during anesthesia. Can J Anaesth 2004; 51: 723–7
26. Suter PM, Fairley B, Isenberg MD. Optimum end-expiratory 44. Lindgren S, Odenstedt H, Olegard C, Sondergaard S, Lundin S,
airway pressure in patients with acute pulmonary failure. N Stenqvist O. Regional lung derecruitment after endotracheal
Engl J Med 1975; 292: 284–9 suction during volume- or pressure-controlled ventilation: a
27. Tusman G, Scandurra A, Bohm SH, Suarez-Sipmann F, study using electric impedance tomography. Intensive Care
Clara F. Model fitting of volumetric capnograms improves Med 2007; 33: 172–80
calculations of airway dead space and slope of phase III. 45. Barber DC. A review of image reconstruction techniques for
J Clin Monit Comput 2009; 23: 197–206 electrical impedance tomography. Med Phys 1989; 16: 162–9
28. Tusman G, Gogniat E, Bohm SH, et al. Reference values for 46. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of
volumetric capnography-derived non-invasive parameters radiation-induced fatal cancer from pediatric CT. AJR Am
in healthy individuals. J Clin Monit Comput 2013; 27: 281–8 J Roentgenol 2001; 176: 289–96
29. Ream RS, Schreiner MS, Neff JD, et al. Volumetric capnogra- 47. Visser KR. Electric properties of flowing blood and impedance
phy in children. Influence of growth on the alveolar plateau cardiography. Ann Biomed Eng 1989; 17: 463–73
slope. Anesthesiology 1995; 82: 64–73 48. Tusman G, Bohm SH, Suarez-Sipmann F, Scandurra A,
30. Fletcher R, Jonson B, Cumming G, Brew J. The concept of Hedenstierna G. Lung recruitment and positive end-expira-
deadspace with special reference to the single breath test tory pressure have different effects on CO2 elimination in
for carbon dioxide. Br J Anaesth 1981; 53: 77–88 healthy and sick lungs. Anesth Analg 2010; 111: 968–77

Handling editor: J. P. Thompson